Patient-Specific Modeling Reveals Stroke-Related Cerebral Hemodynamic Changes in Adults With Sickle Cell Disease
Lara Abdelmohsen, Anyssa Oden, Rhea Soo, Joel Disu, Nahom Mossazghi, Elizabeth Meinert-Spyker, Tamer Ibrahim, Enrico M Novelli, Sossena Wood, Noelia Grande GutiérrezAbstract
Background
Sickle Cell Disease (SCD) is a lifelong genetic blood disorder associated with an increased risk of cerebrovascular disease and stroke, beginning in childhood and persisting into adulthood. Transcranial Doppler (TCD) screening using a velocity threshold >200 cm/s in the internal carotid, middle cerebral, and anterior cerebral arteries of the Circle of Willis has significantly reduced stroke incidence in children with SCD. However, no equivalent evidence-based cerebrovascular risk assessment guidelines exist for adults, partly due to reduced acoustic windows resulting from increased skull thickness. This gap is especially concerning as many adults experience silent or overt strokes without clear early warning signs. Emerging evidence has shown that patient-specific hemodynamic modeling is a useful tool to explore these risks in SCD. This study aims to characterize cerebral hemodynamic patterns in adults with SCD, with and without prior stroke using MRI and computational fluid dynamics, and to evaluate how different approaches of estimating blood flow distribution using CFD influence clinically relevant measures of cerebrovascular health.
Methods
Adult participants (over 18 years old) included healthy controls (n = 3) and adults with SCD without prior ischemic stroke (n = 3) and post-stroke (n = 3). The stroke patients were all male, thus we selected male healthy controls and adults with SCD without prior stroke to match. High-resolution magnetic resonance imaging (MRI) was used to reconstruct patient-specific three-dimensional models of the CoW in SimVascular. Computational simulations were used as a quantitative analysis tool to assess blood flow velocity, pressure differences across the cerebral arteries, and wall shear stress, a measure of the frictional force of blood on vessel walls that has been linked to vascular remodeling and disease. In a separate cohort, one healthy control and one adult with SCD post-stroke were examined to compare three methods of estimating downstream blood flow distribution: (1) Arterial spin labeling (ASL)-based regional perfusion, (2) distribution based on vessel size (area-based), and (3) distribution based on the relative volume of brain vascular territories (volume-based). Blood flow waveforms were measured using phase-contrast MRI, and regional brain perfusion was measured using arterial spin labeling (ASL). Outcomes included pressure drop, mean blood flow, wall shear stress, and velocity.
Results
Differences in cerebral blood flow patterns were observed between groups. Adults with SCD and stroke history demonstrated lower blood flow velocities at TCD-equivalent locations in the CoW (with the exception of one participant with carotid artery stenosis), the lowest average wall shear stress, and the largest vessel surface area exposed to low shear stress (< 1 Pa). In contrast, adults with SCD with no stroke history exhibited higher blood flow velocities, higher wall shear stress, and greater exposure to high shear stress (>7 Pa). Despite similar total cerebral blood flow compared to healthy controls, SCD with stroke history showed a reduced pressure drop across the CoW, suggesting altered vascular regulation. In the blood flow distribution comparison, distributions based on vascular territory volume most closely matched ASL perfusion measurements, with pressure and flow differences typically within 1–14%. Area-based methods (current CFD standard) produced large errors, particularly in the posterior circulation, sometimes exceeding 100% difference in predicted flow. These discrepancies also affected velocity and shear stress patterns, which are commonly used to assess disease progression.
Conclusions
Adults with SCD demonstrate distinct cerebrovascular flow patterns that differ based on stroke history, even when total brain blood flow appears similar. These findings suggest that stroke risk in adults with SCD may be driven not only by how much blood reaches the brain, but how it is distributed across cerebral arteries. Accurate, physiology-informed assessment of cerebral blood flow is therefore critical. This work highlights the potential of advanced brain imaging and computational modeling to support improved stroke risk stratification in adults with SCD and highlights the need for adult-specific cerebrovascular screening and prevention strategies.